Conditions of dementia such as Alzheimer's disease (AD) are frequently characterised by a progressive accumulation of intracellular and/or extracellular deposits of proteinaceous structures such as β-amyloid plaques and neurofibrillary tangles in the brains of affected patients. The appearance of these lesions largely correlates with pathological neurofibrillary degeneration and brain atrophy, as well as with cognitive impairment (Mukaetova-Ladinska, E. B. et al. (2000) Am. J. Pathol. Vol. 157, No. 2, 623-636).
Both neuritic plaques and neurofibrillary tangles contain paired helical filaments (PHFs), of which a major constituent is the microtubule-associated protein tau (Wischik et al. (1988) PNAS USA 85, 4506). Plaques also contain extracellular β-amyloid fibrils derived from the abnormal processing of amyloid precursor protein (APP; Kang et al. (1987) Nature 325, 733). An article by Wischik et al. (in ‘Neurobiology of Alzheimer's Disease’, 2nd Edition (2000) Eds. Dawbarn, D. and Allen, S. J., The Molecular and Cellular Neurobiology Series, Bios Scientific Publishers, Oxford) discusses in detail the putative role of tau protein in the pathogenesis of neurodegenerative dementias.
Studies of Alzheimer's disease indicate that the loss of the normal form of tau (Mukaetova-Ladinska et al. (1993) Am. J. Pathol., 143, 565; Wischik et al. (1995a) Neurobiol. Ageing, 16: 409; Lai et al. (1995b) Neurobiol. Ageing, 16: 433), accumulation of pathological PHFs (Mukaetova-Ladinska et al. (1993), loc. cit.; Harrington et al. (1994a) Dementia, 5, 215; Harrington et al. (1994b) Am. J. Pathol., 145, 1472; Wischik et al., (1995a), loc. cit.) and loss of synapses in the mid-frontal cortex (Terry et al. (1991) Ann. Neurol., 30, 572) correlate with associated cognitive impairment. Furthermore, loss of synapses (Terry et al., loc. cit.) and loss of pyramidal cells (Bondareff et al. (1993) Arch. Gen. Psychiatry, 50: 350) both correlate with morphometric measures of tau-reactive neurofibrillary pathology, which parallels, at a molecular level, an almost total redistribution of the tau protein pool from a soluble to a polymerised form (PHFs) in Alzheimer's disease (Mukaetova-Ladinska et al. (1993), loc. cit.; Lai et al. (1995), loc. cit.).
Tau exists in alternatively-spliced isoforms, which contain three or four copies of a repeat sequence corresponding to the microtubule-binding domain (Goedert, M., et al. (1989) EMBO J. 8, 393-399; Goedert, M., et al. (1989) Neuron 3, 519-526). Tau in PHFs is proteolytically processed to a core domain (Wischik, C. M., et al. (1988) Proc. Natl. Acad. Sci. USA 85, 4884-4888; Wischik et al. PNAS USA 1988, 85:4506-4510); Novak, M., et al. (1993) EMBO J. 12, 365-370) which is composed of a phase-shifted version of the repeat domain; only three repeats are involved in the stable tau-tau interaction (Jakes, R., et al. (1991) EMBO J. 10, 2725-2729) Once formed, PHF-like tau aggregates act as seeds for the further capture and provide a template for proteolytic processing of full-length tau protein (Wischik et al. 1996 Proc Natl Acad Sci USA 93, 11213-11218).
In the course of their formation and accumulation, paired helical filaments (PHFs) first assemble to form amorphous aggregates within the cytoplasm, probably from early tau oligomers which become truncated prior to, or in the course of, PHF assembly (Mena, R., et al. (1995) Acta Neuropathol. 89, 50-56; Mena, R., et al. (1996) Acta Neuropathol. 91, 633-641). These filaments then go on to form classical intracellular neurofibrillary tangles. In this state, the PHFs consist of a core of truncated tau and a fuzzy outer coat containing full-length tau (Wischik., C. M., et al, (1996) loc. cit.). The assembly process is exponential, consuming the cellular pool of normal functional tau and inducing new tau synthesis to make up the deficit (Lai, R. Y. K., et al., (1995), Neurobiology of Ageing, Vol. 16, No. 3, 433-445). Eventually, functional impairment of the neurone progresses to the point of cell death, leaving behind an extracellular tangle. Cell death is highly correlated with the number of extracellular tangles (Wischik et al. 2000, loc.cit). As tangles are extruded into the extracellular space, there is progressive loss of the fuzzy outer coat of the neurone-PHF with corresponding loss of N-terminal tau immunoreactivity, but preservation of tau immunoreactivity associated with the PHF core (FIG. 1; also Bondareff, W. et al., (1994) J. Neuropath. Exper. Neurol., Vol. 53, No. 2, 158-164).
The phase shift which is observed in the repeat domain of tau incorporated into PHFs suggests that the repeat domain undergoes an induced conformational change during incorporation into the filament. During the onset of Alzheimer's disease, it is envisaged that this conformational change could be initiated by the binding of tau to a pathological substrate, such as damaged or mutated membrane proteins (see FIG. 2—also Wischik, C. M., et al. (1997) in “Microtubule-associated proteins: modifications in disease”, eds. Avila, J., Brandt, R. and Kosik, K. S. (Harwood Academic Publishers, Amsterdam) pp.185-241).
In the case of Alzheimer's disease, current pharmaceutical therapies are focused on symptomatic treatment of the loss of cholinergic transmission which results from neurodegeneration (Mayeux, R., et al. (1999) New Eng. J. Med. 341, 1670-1679). However, although the available treatments delay progression of the disease for up to six to twelve months, they do not prevent it. The discovery of drugs that could prevent the aggregation of tau which leads to neurodegeneration would provide a more effective strategy for prophylaxis or for inhibiting the progression of the disease, which would not require an immediate knowledge of the diverse upstream events that initiate the aggregation (see FIG. 3).
Models and Assays
WO 96/30766 describes an in vitro assay for tau aggregation in which a fragment of tau corresponding to the core repeat domain, which has been adsorbed to a solid phase substrate, is able to capture soluble full-length tau and bind tau with high affinity (see FIG. 4). This association confers stability against digestion of proteases on the tau molecules on the repeat domains of tau molecules which have aggregated. The process is self-propagating, and can be blocked selectively by prototype pharmaceutical agents ((Wischik et al. 1996 Proc Natl Acad Sci USA 93, 11213-11218).
Although the in vitro assay described in WO 96/30766 enables the identification of inhibitors or modulators of tau-tau association, the present inventors have also recognized that cell-based models of Alzheimer's disease-like protein aggregation would be useful. Such cellular models could be used both in the primary screening of candidate modulators of tau-tau aggregation, and in the secondary screening of compounds already identified in the in vitro assay of WO 96/30766. Furthermore, the demonstration of tau aggregation in cells could also aid in the identification of normal cellular substrates which are involved in the initiation of pathological tau aggregation, which substrates could themselves be targets for pharmaceutical intervention.
However, numerous papers reporting the expression of various tau constructs in tissue culture models have failed to demonstrate aggregation (see e.g. Baum, L. et al., (1995) Mol. Brain Res. 34:1-17). For instance, 3T3 mouse fibroblasts do not possess tau protein and thus present a cellular environment in which recombinant tau can be expressed independent of endogenous mouse tau. Transfection of various cell lines has been reported previously (Kanai et al., 1989; Goedert and Jakes, 1990; Knops et al, 1991; Lee and Rook, 1992; Gallo et al., 1992; Lo et al., 1993; Montejo de Garcini et al., 1994; Fasulo et al., 1996). However the stable long term expression of truncated tau in such cell lines was not achieved. For example, tau constructs for residues 164 or 173 to 338 or 352 did not express protein (Lee and Rook, 1992).
Although Fasulo et al. (Alzheimer's Research 1996, 2, 195-200) reported transient expression of truncated tau in COS cells, data for stable long term expression of this tau was not shown. These workers concluded from the use of the transient transfection system that expression of truncated tau by itself was not sufficient to induce tau aggregation in a manner suitable for testing drugs.
Thus far, the aggregation of soluble tau in vitro has only been achieved under non-physiological conditions and at high concentrations (reviewed in Wischik (2000), loc. cit).
WO 96/30766 describes two approaches for studying tau aggregation in a cellular environment. In the first approach, full-length tau or fragments of tau were stably expressed in cells. In the second approach, aggregated tau was transiently transfected into cells by use of lipofectin.
Although both of these approaches are useful for the study of tau-tau aggregation, they have some limitations. Transfection of aggregated tau into cells using lipofection is of variable efficiency, as is the production in vitro of aggregated tau itself. Moreover, the core tau fragment, which is the most efficient seed for tau aggregation, is found to be toxic when stably expressed in cells, leading to low expression levels. Thus, constitutive expression of the truncated tau fragment of the PHF core in eukaryotic cells is difficult to achieve. Transient expression systems permit the optimization of expression of tau, but the inherent toxicity of the fragments renders even these systems unreliable. Longer fragments of tau are less toxic, but these do not reliably aggregate when expressed in cells.
Thus it would be desirable for an alternative model system to be developed, in which the interaction between e.g. tau molecules and the like could be investigated under physiological conditions, in a stable and controllable cell line, and which could be used to screen for potential diagnostic, prognostic or therapeutic agents of conditions such as Alzheimer's disease.